by Carl Sagan
The actual discovery of the moons of Mars was made from outside Washington, D.C., in 1877. The U. S. Naval Observatory had just completed a large refracting telescope. The observatory’s astronomer, Asaph Hall, sought to learn whether the moons of Mars, fabled in song and story, actually existed. His first few nights were unsuccessful, and, in a despondent mood, he announced to his wife his intention of quitting the search. Mrs. Hall would have none of this, and urged her reluctant husband back to a few more nights at the telescope–whence he emerged with two Martian satellites in hand. For a brief period he thought he had found three satellites, because the inner one moved so rapidly that on one night he saw it on one side of Mars and on the next night on the other. Hall named the moons Phobos and Deimos, after the horses that pulled the chariot of the god of war in Greek mythology; they have the cheerful denotations of fear and terror, respectively. (The appropriate adjectives pose some problems: Do we talk of Phobic orbits and Deimonic nights?) If another moon of Mars is someday found, I hope we will give it a less ferocious and more optimistic name–like “Pax.”
I also hope that when the features on Phobos and Deimos are eventually named by the International Astronomical Union, one will be named after Mrs. Hall. But since another feature will surely be named after Asaph Hall, we have a problem–two craters named Hall would be confusing. In an astronomy talk I gave at Harvard University, I commented wistfully that the problem would be solved if only we knew Mrs. Hall’s maiden name. My friend Owen Gingerich, Professor of the History of Science at Harvard, instantly leaped to his feet with the words “Angelina Stickney” tripping off his lips. So when the time comes, I hope there will be a “Stickney” on one of the moons of Barsoom.
The subsequent study of Phobos and Deimos between 1877 and 1971 has a curious history. The moons of Mars are so tiny that they appear, even to the largest Earthbound telescopes, as dim points of light. They are too faint for the pre-1877 telescopes to have seen them at all. Their orbits can be calculated by noting their positions at various times. In 1944 at the U. S. Naval Observatory (where an understandably proprietary interest in Phobos and Deimos must have developed), B. P. Sharpless collected all the observations available in his day to determine the orbits to the best possible precision. He found–no doubt to his surprise–that the orbit of Phobos appeared to be decaying, what astronomers call a secular acceleration. Over longish periods of time the satellite seemed to be approaching more and more closely to Mars and moving more and more rapidly. This phenomenon is quite familiar to us today. The orbits of artificial satellites are decaying all the time in the Earth’s atmosphere. They are initially slowed by collisions with the diffuse upper atmosphere of the Earth, but by Kepler’s laws the net result is a more rapid motion.
Sharpless’ conclusion of a secular acceleration for Phobos remained an unexplained and almost unexamined curiosity until it was considered around 1960 by the Soviet astrophysicist I. S. Shklovskii. Shklovskii considered a wide range of alternative hypotheses for the secular acceleration, among them the influence of the Sun, the influence of a hypothetical magnetic field on Mars, and the tidal influence of the gravity of Mars. He found that none of these came close to working. He then reconsidered the possibility of atmospheric drag. The exact size of the Martian satellites was known poorly and indirectly in those days before spacecraft investigation of Mars, but it was known that Phobos was roughly ten miles in diameter. The altitude of Phobos above the surface of Mars was also known. Shklovskii and others before him found that the density of the atmosphere was far too low to produce the drag that Sharpless had deduced. It was at this point that Shklovskii made a brilliant and daring guess.
All the calculations showing atmospheric drag to be ineffectual had assumed that Phobos was an object of ordinary density. But what if its density were very low? Despite its enormous size, its mass would then be quite small, and its orbit could be appropriately affected by the thin upper atmosphere of Mars.
Shklovskii calculated the required density of Phobos, and found a value about one one-thousandth the density of water. There is no natural object or substance with such low density; balsa wood, for example, has about half the density of water. With such a low density, there was only one conclusion possible: Phobos had to be hollow. A vast hollow object ten miles across could not have arisen by natural processes. Shklovskii, therefore, concluded that it was produced by an advanced Martian civilization. Indeed, an artificial satellite ten miles across requires a technology far in advance of our own; it would also be far in advance of the technology imagined on Barsoom by Burroughs, which was a kind of sword and small spaceship technology.
Since there were no signs of such an advanced civilization on Mars today, Shklovskii concluded that Phobos–and possibly Deimos–had been launched in the distant past by a now extinct Martian civilization. (The interested reader may find more details of this remarkable argument of Shklovskii’s in the book Intelligent Life in the Universe, jointly authored by Shklovskii and myself. [San Francisco, Holden-Day, 1966; New York, Delta Books, 1967]). Subsequent to Shklovskii’s first work on the subject, the motions of the moons of Mars were reexamined in England by G. A. Wilkins, who found that possibly there was no secular acceleration. But he could not be sure.
Shklovskii’s extraordinary suggestion that the moons of Mars might be artificial is one of three hypotheses on their origin. The other two–certainly interesting in their own right, but naturally paling in comparison with the Shklovskii hypothesis–are (1) that the moons are captured asteroids, or (2) that they are debris left over from the origin of Mars itself.
Asteroids are hunks of rock and metal that go around the Sun between the orbits of Mars and Jupiter. There are unlikely, but theoretically possible, scenarios in which the gravitation of Mars can capture a close-passing asteroid.
In the Martian-debris hypothesis, it is imagined that pieces of rock of various sizes fell together to form Mars; that the last generation of such infalling pieces produced the large old impact craters on Mars (see page 131); and that Phobos and Deimos are by chance the only remnants still extant of the early catastrophic history of Mars.
It is clear that establishing any one of these three hypotheses on the origin of the moons of Mars would be a major scientific achievement.
The Mariner Mars mission of 1971, which I had the pleasure to work on, was originally to have involved two spacecraft, Mariner 8 and Mariner 9. They were to be placed in different orbits for different purposes in the study of Mars itself. After these orbits were finally agreed upon, I noticed that they were not all that far from the orbits of Phobos and Deimos. It also seemed to me that television and other close-up observations of Phobos and Deimos by the Mariner spacecraft might permit us to determine something of their origin and nature.
I therefore approached officials of NASA, which organized and ran the mission, for permission to program observations of Phobos and Deimos. While the mission controllers at Jet Propulsion Laboratory, the actual operating organization, were not unsympathetic to this idea, some officials at NASA headquarters were against it. There was a mission plan, written in a large book, which stated what Mariners 8 and 9 were about. Nowhere in the mission plan were Phobos and Deimos mentioned. Ergo, I could not look at Phobos and Deimos.
I pointed out that my proposal required only moving the scan platforms on the spacecraft so that the cameras could observe the Martian satellites. The response was negative again. A short time later, I advanced the argument that if Phobos and Deimos were indeed captured asteroids, examining them from Mariner 9 was the equivalent of a free mission to the asteroid belt: The proposed scan platform maneuver would save NASA two hundred million dollars or so. This argument was judged, at least in some circles, to be more compelling. After about a year of my lobbying, a planning group on satellite astronomy was set up, and tentative plans were made for examining Phobos and Deimos. The satellite astronomy working group was, at my suggestion, chaired by Dr. James Pollack, a former student of mine; but
it was a sign of NASA reluctance that the group was formed only after the launch of Mariner 9, and only about two months before its arrival at Mars. (Mariner 8 had, meanwhile, failed.)
When Mariner 9 arrived at Mars, we found a planet almost entirely obscured by dust. Since there was little to look at on Mars, a great and previously undetectable enthusiasm for examining Phobos and Deimos dramatically materialized. The first step was to take wide-angle photographs from a distance in order to establish with some precision the orbits and locations of the moons. This task was accomplished in a preliminary way about two weeks after injection of the spacecraft into Martian orbit. Mariner 9 has an orbital period of about twelve hours, so that it made close to two revolutions around Mars per day.
The television pictures from Mariner 9 were radioed from Mars to Earth in much the same way that a newsprint wire-photo is transmitted on Earth. The picture is divided into a large number of small dots (for Mariner 9, several hundred thousand dots), each dot with its own brightness, or shade of gray, running from black to white. After the picture is taken by the spacecraft and recorded there on magnetic tape, it is played back to Earth, dot by dot. The communication says, in effect: Dot number 3277, gray level 65; dot number 3278, gray level 62, and so on. The picture is reassembled by computer on the Earth–essentially by following the dots.
The first moderately close-up photograph of Phobos was obtained on revolution 31. Page 100 shows a Polaroid photo of the video-monitor image of Phobos on revolution 31, received on November 30, 1971. The image is much too indistinct to make any conclusions whatever.
Late that same night, Dr. Joseph Veverka, of Cornell, another former student of mine, and I worked into the small hours of the morning at the Image Processing Laboratory of JPL to bring out–by computer contrast-enhancement techniques–all of the detail present in the image. The result is shown on page 102. The shape is irregular. Are those blotches craters?
Our computer-enhanced photograph was constructed on the computer’s video monitor, line by line, from top to bottom. As the apparent large crater at the top gradually emerged, we saw a single bright spot at its center; for just a moment, I had the sense that we were seeing a star through an enormous hole in Phobos–or, even more chilling, that we were seeing an artificial light. But when we requested the computer to remove all single-bit errors, the bright spot went away.
On revolution 34, Mariner 9 and Phobos came within less than four thousand miles of each other, one of the closest approaches in the entire mission. Late on the night of the receipt of that picture, Veverka and I were again computerenhancing. Our results were like those seen on the accompanying page 103. I am not sure what an artificial satellite ten miles across looks like, but this does not seem to be it. Phobos looks not so much like an artificial satellite as a diseased potato. It is, in fact, very heavily cratered. For it to have accumulated so many craters in that part of the Solar System, it must be very old, probably billions of years old.
Phobos appears to be an entirely natural fragment of a larger rock severely battered by repeated collisions; holes have been dug, pieces have been chipped off. It looks a little like the hand axes, chipped along natural fracture planes, made by our Pleistocene ancestors. There is no sign of technology on it. Phobos is not an artificial satellite. When pictures of Deimos were computer contrast-enhanced, the same conclusion applied to it.
Phobos and Deimos are the first satellites of another planet to have been photographed close-up. They were also observed by the ultraviolet spectrometer and the infrared radiometer on board Mariner 9. We have been able to determine their sizes and shapes and something of their color. They are extremely dark objects–darker than the darkest material that is likely to be in the room in which you are sitting right now.
Indeed, they are among the very darkest objects in the Solar System. Since there are so few objects this dark anywhere, we hope to be able to conclude something about their composition. They are both covered by at least thin layers of finely pulverized material. They provide important clues to collisional processes in the early Solar System. I believe we are looking at the end-product of a kind of collisional natural selection, in which fragments have been broken off from a larger parent body, and we are seeing only the two pieces, Phobos and Deimos, that remain. The moons of Mars are also important collision calibrators for Mars. Phobos, Deimos, and Mars have very likely been together in the same part of the Solar System for a very long period of time. The number of craters of a given size on Mars is much less, in general, than on Phobos and Deimos, providing important information on erosional processes that exist on Mars and that do not exist on airless and waterless Phobos and Deimos.
Because we now have the first good information on the size and shape of these objects, and because we now have good reason to think that they have typical densities of ordinary rock, we can calculate something about what it would be like to stand on, let’s say, Phobos. First of all, Mars, less than six thousand miles away, would fill about half the sky of Phobos. Marsrise would be a spectacular event. Eventual construction of an observatory on Phobos to examine Mars might not be such a bad idea. We know from Mariner 9 that both Phobos and Deimos are rotating as our Moon does, always keeping the same face to their planet. When Phobos is above the day hemisphere of Mars, the reddish light of Mars would be enough to read by at night on Phobos.
Because of their small sizes, Phobos and Deimos have very low gravitational accelerations. Their gravities do not pull very hard. The pull on Phobos is only about one one-thousandth of that on Earth. If you can perform a standing high jump of two or three feet on Earth, you could perform a standing high jump of half a mile on Phobos. It would not take many such jumps to circumnavigate Phobos. They would be graceful, slow, arcing leaps, taking many minutes to reach the high point of the self-propelled trajectory and then to return gently to the ground.
Even more interesting would be a game like baseball on Phobos. The velocity necessary to launch an object into orbit about Phobos is only about twenty miles per hour. An amateur baseball pitcher could easily launch a baseball into orbit around Phobos. The escape velocity from Phobos is only about thirty miles per hour, a speed easily reached by professional baseball pitchers. A baseball that had escaped from Phobos would still be in orbit about Mars–a man-launched moonlet. If Phobos were perfectly spherical, a lonely astronaut with an interest in baseball could invent a curious but somewhat sluggish version of this already rather sluggish game. First, as pitcher, he could throw the ball sidearm–at the horizon at between twenty and thirty miles per hour. He could then go home for lunch, because it will take about two hours for the baseball to circumnavigate Phobos. After lunch, he can pick up a bat, face the other direction and await his pitch of two hours earlier. Apart from the fact that good pitchers are seldom good hitters, hitting this pitch would be pretty easy: About fifteen seconds elapse from the appearance of the baseball at the horizon to its arrival in the vicinity of our astronaut. If he swings and misses–or, more likely, if the ball is wide of the plate–he can then go home for a two-hour nap, returning with his catcher’s mitt to catch the ball. Alternatively, if he succeeds in hitting a fly ball at a velocity somewhere between twenty and thirty miles per hour, he can go home and take his nap, returning this time with a fielder’s mitt, awaiting the return of the ball from the opposite horizon two hours later. Because Phobos is gravitationally lumpy, the game would be more difficult than I have indicated. Since daylight on Phobos lasts only about four hours, lights would have to be erected, or the game modified so that all pitching, hitting, and catching events happen on the day side.
These sports possibilities may, one day a century or two hence, provide a tourist industry for Phobos and Deimos. But baseball on Phobos is no more an argument for going there than, to take a random example, golf is for going to the Moon. The scientific interest in the moons of Mars–whether captured asteroids or debris from the formation of the planet–is, however, immense. Sooner or later, certainly on a time s
cale of centuries, there will be instruments–and then men–on the surface of Phobos looking up with awe at an immense red planet that fills the sky from zenith to horizon.
And what about the opposite view? What do the moons of Barsoom look like from the surface of Mars? Because Phobos is so close to Mars, it would be seen as a clearly discernible disc, even though it is intrinsically such a tiny object. In fact, Phobos would appear as about half the apparent size of our Moon seen from the surface of Earth. We have found from Mariner 9 that only one side of Phobos is visible from Mars, just as only one side of our Moon is visible from Earth. That face of Phobos is, more or less, the face on page 102. Until Mariner 9, no one–except Martians, if such there be–ever knew that face.
Because Phobos is so close to Mars, Kepler’s laws constrain it to move comparatively rapidly about the planet. It makes approximately 2½ revolutions about Mars in 24 hours. Deimos, on the other hand, takes 30 hours 18 minutes to revolve in its orbit once about Mars. Both moons revolve in their orbits in the same direction or sense as Mars rotates on its axis. Thus, Deimos rises in the east and sets in the west as–from terrestrial chauvinism–we believe a well-behaved satellite should. But Phobos makes it once around its orbit in less time than it takes for Mars to rotate. Accordingly, Phobos rises in the west and sets in the east, taking about 5½ hours to transit from horizon to horizon. This is not exactly “hurtling”–the motion would not be easily perceptible against the field of stars in a minute’s watching–but it’s not plodding, either. There will be some nights at the equator on Mars when Phobos sets in the east at sunset and then rises in the west well before dawn.